Writing about 3-D is a bit like trying to do play-by-play commentary on a hockey game being played at twice normal speed. By the time you comment on one play, the game has moved on to such a degree that it's already irrelevant. Maybe it's not quite that bad, but it is a challenge to talk about a topic that's moving at unprecedented speed, while trying to ensure your comments remain relevant.
The main purpose of this article is to help you consider your state of readiness for 3-D TV. Even if you don't think this is going to affect you any time soon, it will be helpful to have an appreciation for what it will take if and when your boss suddenly comes to you and says, “Tell me right now what it would take for us to hop on this 3-D bandwagon!”
Brief background
3-D TV can be characterized as the distribution of two signals (left eye and right eye) to the viewer, so that a stereoscopic view of the content is achieved. The current 3-D movement comes, naturally, out of cinema. The reason it's different from previous, abandoned attempts at 3-D is largely due to cinema's (and TV's) move to digital. Today's 3-D is of far superior quality, less prone to audience discomfort and seems much more likely to succeed than previous attempts.
With the clear ROI for 3-D releases demonstrated by studios, it's logical that this movement would get the attention of those producing and distributing content to TV viewers. Add to that the release of 3-D Blu-ray players and titles to consumers, upcoming cable and satellite 3-D offerings, and the onslaught of 3-D TVs — many at relatively low incremental cost to consumers over traditional HDTVs — and you have something that seems to have many of the ingredients for success.
Content drivers
Those in the TV business who specialize in sports and movie content are the first to embrace 3-D TV. Big sporting events and feature films have appeal in 3-D that goes well beyond other TV fare. For viewers, donning glasses to watch special events like these seems reasonable, but expecting them to put on and take off glasses for everything under the sun may be unrealistic. So, if you specialize in news and typical TV series, you may not have to worry as much about 3-D content today as others will, but never say never.
Let's clear up some confusion
One of the areas of frustration for broadcasters at this point is the confusion surrounding the various 3-D technologies. A quick survey of that landscape shows why 3-D can be bewildering.
The 2-D + Difference approach, which has enjoyed success in gaming and other areas, involves taking a traditional 2-D picture and adding metadata describing the difference (or depth) between the left eye and right eye view to allow TVs to render the second eye view. Blu-ray is going with MPEG's MVC (Multiview Video Coding) extension of H.264/MPEG-4 AVC, which supplements a traditional 2-D view with metadata, enabling the construction of the second eye view. This provides full-resolution pictures for each eye. Of course, bandwidth isn't as much of a consideration for Blu-ray as it is for other distribution media.
Frame-compatible approaches, such as over-under and side-by-side are favored by cable and satellite providers. There are others, such as line interleave, column-interleave and checkerboard, but those don't seem to be enjoying the uptake of the other two. Frame-compatible approaches make sense for distribution media with limited bandwidth (our current HD distribution pipes).
Frame-sequential approaches, which are based on every other frame being targeted at a different eye, offer full HD resolution — but at a cost. This is the format of choice for lower-cost displays, utilizing active shutter glasses, but doesn't make sense for early distribution of broadcast material.
An easy trap to find yourself in if you're a TV person trying to sort through all of this is a tendency to confuse the transmission and display formats with the formats used to handle 3-D internally in the plant. The format used in production, post production and processing in your facility need not match that used in transmission or viewing.
Here's an example. The viewer may watch the content with a frame-sequential display running at 120Hz (or faster) with active shutter glasses. The signal may reach him via satellite in a frame-compatible, side-by-side format. However, the program may have been recorded as two distinct full HD streams and handled internally in the broadcast plant as two streams. Decoding one 3-D format and encoding it into another is just as routine as moving from any 2-D format to another, so be careful to decouple the emission and display format used from that applied in the production and transmission facilities.
Since many are still reeling from a recent investment in a digital infrastructure, expecting them to undergo another plant upgrade — replacing all the associated gear with special 3-D gear — is unrealistic in many cases. It is for this same reason that cable and satellite operators currently favor frame-compatible 3-D TV approaches. They may not be full HD for each eye (as they must drop half the resolution, either vertically or horizontally, to fit two views into a single frame), but these approaches allow stereoscopic content to be handled by existing infrastructure with mostly minor upgrades. And reduced resolution or not, it looks pretty darned good.
However, these frame-compatible approaches seem unlikely to be adopted internally by broadcasters, as lost resolution in the plants is something most will try to avoid. Alternatives, such as 2-D + Difference, which is now touted as requiring only a 35 percent increase in bandwidth and MVC, seem to be interesting approaches to dealing with identical 2-D and 3-D content (since 2-D + Difference and MVC are 2-D- and 3-D-compatible in a single stream and full resolution for each eye). At the moment, however, it would seem more likely that they'll instead use two full HD streams internally.
Practically, we should think about 3-D TV in a similar fashion to SD and HD signals — with just a few exceptions. Figure 1 shows a typical 3-D workflow.
Post production
There is no question that preserving the two streams in the highest quality possible is most important — mezzanine I-frame compression only. As for monitoring the use of 3-D TV content, active shutter glasses should be used with full-bandwidth video into the monitor.
In plant
Ideally, 3-D TV video within a broadcast facility will match up with the HD plant format. The left eye and right eye streams will be 1080i or 720p on a single 3Gb/s or two 1.5Gb/s coax. Certainly running on a single 3Gb/s coax has several advantages, similar to the change from mono into stereo audio (matched timing, switching and processing).
Compression for storage is a touchy subject. Longer GOP compression is being used more and more for HD quite successfully. There are always compression artifacts; however, in 3-D TV, we need to ensure that the artifacts are the same on both the left and right channels. Practically speaking, the only way to ensure this is to use higher compression bit rates than typically used for HD.
Usually, several channels are run together within a facility, and this makes keeping an eye on several monitors on the same wall difficult to impossible using active shuttered glasses. Monitor walls and multiviewers are best suited for passive polarized glasses.
Contribution
Whether the 3-D TV content is backhauled from an event or being sent out to affiliates, preserving the left and right eye channel is of prime importance. Naturally, dark fiber is the best option, running at 3Gb/s. The next best options are using two streams with a synchronized MPEG higher bit rate encoding process or the higher bit 2-D + Difference approach mentioned earlier.
Distribution
Ultimately, we will see 2-D + Difference or some other full-resolution approach to 3-D for distribution in the long term. As long as the bandwidth concerns can be dealt with, it would seem logical that most broadcasters and consumers would, given a choice, prefer full HD for each eye.
For now, however, that isn't going to happen, as current cable and satellite paths and STBs will be used. The best approach is squeezing the video by half and putting left and right channel adjacent within the same stream. This is most commonly done either side by side, or one above the other. As this preserves the frame (and field) nature of the video, it can go through today's compression engines and links generating the same artifacts on both channels. The aforementioned details are summarized in Table 1.
Are you ready now?
We hope this article increased your awareness of some of the technologies being used for 3-D today, as well as into the future. We also hope it cleared up some misconceptions and confusion about 3-D, and its potential impact on your plant.
The main lesson here is that a 3Gb/s infrastructure will be a great investment for your future ability to handle 3-D video. Talk with your primary vendors and inquire about the readiness of the products you use to handle 3-D. In some cases, they may be ready now. In others, an upgrade may be needed. Replacement of equipment won't be required in many cases.
Key takeaways on 3-D TV? Understand the technology, talk to your vendors, and be ready for that day when the boss walks in and asks the dreaded 3-D question … !
Stan Moote is vice president of corporate development and Chris Lennon, CTO Group, for Harris Broadcast Communications.
Table 1. Format, compression and display types vary at different stages in 3-D workflow. Production In plant Contribution Distribution Format Dual stream Dual stream Dual stream or 2-D + Difference Half resolution (adjacent) Compression Mezzanine Mid-high bit rates Higher backhaul rates Today's HD to the home bit rates Display types Active shutter glasses Passive polarized glasses Active shutter glasses (home)
What action is the standards world taking?
Just as it is across the media industry in general, 3-D is a big deal these days in virtually every related standards development organization. Different organizations have different perspectives and plays in the world of 3-D, and this is reflected in their current activities.
A few of interest include:
- 3D@Home ConsortiumThis organization has several steering teams, focusing on everything from creation, all the way to display of 3-D content.
- ATSCAlthough 3-D initially fell within the scope of future work, at least one ATSC country is expecting to hold trials of 3-D over-the-air broadcasting later this year.
- CEAAs consumer electronics manufacturers roll out a wide array of 3-D TV equipment (including Blu-ray players), the CEA has been a leader in this area. It has led several consumer surveys and has a 3D Task Force under way.
- DVBWith a new group recently formed, DVB is looking at 3-D potential for its constituent countries. We can expect more news here very soon.
- ITUThis organization is looking far beyond the stereoscopic 3-D that much of the industry is now focused on. It considers the current work to be only the first of a three-phased implementation of 3-D TV, culminating in holographic television.
- MPEG IFIts new 3DTV Working Group is studying some interesting topics, including support of 2-D/3-D ad insertion.
- SCTESCTE formed a 3D Working Group that has been quite active for some time already, and is working closely with CableLabs on cable-specific considerations in the area of 3-D.
- SMPTEFollowing publication of its 3D Task Force report, SMPTE has working groups studying issues in the following areas: 3-D image format, graphics, subtitles and metadata. Output from these groups is anticipated in mid-2010.
